An Intelligent Real Time Design Methodology for Component Selection: An Approach to Managing Uncertainty

Abstract

An Intelligent Real Time Design (IRTD) methodology is presented for component selection applications under the reality of uncertain and incomplete information. A decision analytic approach is developed with the goal of assisting designers in making decisions that balance the cost of the limited resources consumed during the design process, such as the designer’s time, against the benefit to be derived from the utilization of those resources in terms of expectations of an improved design. This approach is shown to complement other formal methods for design based on interval arithmetic and qualitative optimization, and a general methodology is proposed for performing component selection utilizing a combination of these methods. An example application to the selection of rolling element bearings is presented to clarify the methodology and demonstrate its effectiveness in providing guidance to designers when selecting elements from a component database.

Product Attribute Function Deployment for Attribute Identification, Concept Selection, and Target Setting

“…Additionally, X can often be represented by discrete values in the conceptual design phase, for example representing catalog component options [2]. The maximization of utility can be evaluated using a genetic algorithm which is commonly used in conceptual design selection when a combination of discrete and continuous design variables are present [6].…”

Model refinement decisions using the process performance indicator

“…From the decision-centric perspective (Mistree et al 1989, Mistree et al 1992 adopted in this article, model refinement is driven by improved decision-making in the context of a given design problem formulated in terms of often interacting decisions and analysis models. Managing complexity with respect to models may thus involve refining (i) interactions between coupled decisions and analysis models, addressed in detail in the literature (Bradley and Agogino 1994, Budde 1999, Eppinger and Salminen 2001, Panchal et al 2008 individual analysis models. The focus in this article is on the stepwise simulation-based refinement of analysis models that can be formulated mathematically and solved computationally.…”

“…To overcome the assumption of having a perfect model available in the context of model refinement, Panchal et al (2008) present a VoI-based metric called 'improvement potential' to account for the non-probabilistic uncertainty resulting from model refinement or simplification of interactions, as proposed in the literature (Bradley and Agogino 1994, Buede 1999, Eppinger and Salminen 2001. However, the improvement potential metric (Panchal et al 2008) is based on the assumption that information about bounds on the output of the model (i.e.…”

Simulation Model Refinement for Decision Making Via a Value-of-Information Based Metric

“…In the context of engineering design, Bradley and Agogino [4] use this value-of-information metric for a catalog selection problem, where a designer is faced with the task of choosing components from a catalog in order to satisfy some functional requirements. During the conceptual design phase, selection decisions need to be made under significant uncertainty due to limited understanding of requirements and constraints, inability to specify part dimensions, uncertainty in the environmental conditions, etc.…”